Your Postgres service is crash-looping because the container image became unavailable as a result of the recent service disruption that affected our Image Registry. Each restart attempt fails before Postgres can start. Your last clean database checkpoint was at 22:12 UTC on May 19, so your data volume should be intact. To recover, try creating a new deployment by clicking the three-dot menu on your Postgres service, selecting "Settings," and triggering a redeploy of the source image - this will pull a fresh copy of the image. If that does not resolve it, we can look into this further on our end.

I have this identical problem. Will attempt the same solution, after hobby plan builds are restored.

The original poster's Postgres service has recovered and is now running successfully. For the second user with the identical problem, the major service disruption that caused these Postgres crash-loops is now resolved, but there is a separate ongoing incident temporarily pausing builds on free, trial, and hobby plans due to a build queue backlog. Redeploying a Postgres service requires pulling the container image, which may be affected while the Image Registry is degraded. We recommend attempting the redeploy once this second incident is resolved.

This is a solution, redeploy Database image and redeploy backend